Modulation of Basophils' Degranulation and Allergy-Related Enzymes by Monomeric and Dimeric Naphthoquinones

Allergic disorders are characterized by an abnormal immune response towards non-infectious substances, being associated with life quality reduction and potential life-threatening reactions. The increasing prevalence of allergic disorders demands for new and effective anti-allergic treatments. Here we test the anti-allergic potential of monomeric (juglone, menadione, naphthazarin, plumbagin) and dimeric (diospyrin and diosquinone) naphthoquinones. Inhibition of RBL-2H3 rat basophils' degranulation by naphthoquinones was assessed using two complementary stimuli: IgE/antigen and calcium ionophore A23187. Additionally, we tested for the inhibition of leukotrienes production in IgE/antigen-stimulated cells, and studied hyaluronidase and lipoxidase inhibition by naphthoquinones in cell-free assays. Naphthazarin (0.1 µM) decreased degranulation induced by IgE/antigen but not A23187, suggesting a mechanism upstream of the calcium increase, unlike diospyrin (10 µM) that reduced degranulation in A23187-stimulated cells. Naphthoquinones were weak hyaluronidase inhibitors, but all inhibited soybean lipoxidase with the most lipophilic diospyrin, diosquinone and menadione being the most potent, thus suggesting a mechanism of competition with natural lipophilic substrates. Menadione was the only naphthoquinone reducing leukotriene C4 production, with a maximal effect at 5 µM. This work expands the current knowledge on the biological properties of naphthoquinones, highlighting naphthazarin, diospyrin and menadione as potential lead compounds for structural modification in the process of improving and developing novel anti-allergic drugs.


Introduction
Allergy is an abnormal immune response against non-infectious environmental substances, named allergens [1]. Allergy comprises chronic disorders associated with reduced quality of life, such as eczema or allergic rhinitis, and potential life-threatening reactions, including anaphylaxis and severe asthma episodes [2]. The prevalence of allergic disorders has been increasing globally, affecting roughly 25% of people in developed countries. This increased prevalence has been associated to environmental changes, such as air pollution and ambient temperature increment, which may induce early springs with increased airborne pollen [1]. On the other hand, the ''hygiene hypothesis'' suggests that reduced exposure to microorganisms in early life contributes to an immune system more susceptible to allergic and autoimmune diseases [3]. In the allergic process, immune cells, such as mastocytes, eosinophils, basophils and macrophages, release several mediators (including histamine and leukotrienes) that are responsible for allergic symptoms [4]. Additionally, these mediators may promote the development of different diseases, by inducing pathophysiological changes in the affected organs [1,5]. A classic example is the role of leukotrienes in the pathogenesis of asthma and allergic rhinitis, by inducing bronchoconstriction and increased vascular permeability [6]. Thus, the increased allergy prevalence, together with the deleterious consequences of repetitive exposure to allergens, stresses the need for new strategies to induce immunological tolerance to allergens as well as new antiallergic drugs [1].
Nature continues to be a rich source of novel bioactive molecules, and several plant extracts have been probed for antiallergic properties. Namely, the grape seed extract of Vitis vinifera L. [7], the rhizomes extract of Dioscorea membranacea Pierre ex Prain & Burkill, in which the main active compound was a quinone (dioscoreanone) [8], or the leaf extract of Rhinacanthus nasutus Kuntze, which is rich in naphthoquinones [9]. Naphthoquinones are compounds constituted by two carbonyl groups in a naphthalene skeleton, naturally occurring in plants, fungi, bacteria and lichens, where they playing key survival roles, namely in defence against pathogens [10]. The high biological potential of naphthoquinones has been used in the search of new drugs, such as new anti-allergic drugs. In fact, 1,4-naphthoquinones isolated from R. nasutus were capable of inhibiting RBL-2H3 basophils' degranulation in the micromolar range, and decreasing tumour necrosis factor (TNF)-a and interleukin production [9]. Further studies, with synthetic naphthoquinones, support their anti-allergic properties: 2-alkyl/arylcarboxamido derivatives of 3-chloro-1,4naphthoquinone inhibited the degranulation on mastocytes stimulated with compound 48/80 [11]. On the other hand, allergic reactions are common after temporary tattoos with henna (derived from Lawsonia inermis L.), where lawsone (2-hydroxy-1,4naphthoquinone) is the main compound responsible for dye properties. Still, allergic reactions to henna have been attributed only to the occasional additive p-phenylenediamine [12,13].

Chemicals and reagents
Medium, buffers and supplements for cell culture, including Earle's Balanced Salt Solution (EBSS) were from Gibco, Invitrogen TM (Grand Island, NY, USA) and bovine albumin fraction V solution 7.5% (BSA) was from Sigma-Aldrich (St. Louis, MO, USA).
Quercetin was used as positive anti-degranulation control [20] and the anti-degranulation effects of diospyrin, diosquinone, juglone, menadione, naphthazarin and plumbagin ( Fig. 1), were studied at non-toxic concentrations (determined by testing several concentrations of each compound and using the MTT assay to evaluate the effect on cell viability). The concentrations of the tested compounds in the degranulation assays with different stimuli were kept constant [0.1 mM (NTZ), 1 mM (DQN and PLB), 5 mM (MND) and 10 mM (JGL)], except for diospyrin and quercetin, where a 10 fold higher concentration was also tested in the A23187 assay.
A23187, quercetin and naphthoquinones stocks were dissolved in dimethyl sulfoxide (DMSO), aliquoted and stored at 220uC. We determined the maximal non-interfering solvent concentrations ( Fig. 3; 0.1% and 0.5% DMSO for IgE/antigen and A23187 assays, respectively), as this was a limiting factor for testing higher naphthoquinone concentrations.
IgE/antigen assay. When the IgE/antigen was used as stimulus, cells were incubated during 16 h with 100 ng/ml IgE anti-DNP and with individual naphthoquinones diluted in culture medium. After washing twice with Dulbecco's phosphate buffered saline (DPBS), cells were treated for 1 h, at 37uC, with 100 ng/ml DNP-BSA and with individual naphthoquinones diluted in EBSS supplemented with 0.1% BSA ( Fig. 2A) [21]. After treatments, supernatants were collected in order to quantify released b-hexosaminidase and released histamine, while cell viability assay was performed on adherent cells.
A23187 assay. Before treatment with A23187, cells were incubated with individual naphthoquinones during 15 min, at 37uC. After that, A23187 1 mM was added and cells incubated for 30 min, at 37uC (Fig. 2B). Compounds were freshly diluted prior to cell exposure using EBSS supplemented with 0.1% BSA [21]. bhexosaminidase and histamine release was quantified in supernatants, whereas the MTT cell viability assay was performed on adherent cells.
Cell viability. Cell viability was assessed by the cellular dehydrogenases' dependent reduction of MTT to formazan, which was quantified by the measurement of optical density at 550 nm using a microplate reader (Multiskan ASCENT Ther-moH), as described before [14].
Released b -hexosaminidase quantification. The release of b-hexosaminidase from stimulated-RBL-2H3 cells was measured as previously described [21]. In a 96-wells plate, 50 ml of substrate solution [p-nitrophenyl N-acetyl-D-glucosamine 1.3 mg/ ml in citrate buffer (pH 4.5)] were added to 30 ml of supernatant. The plate was incubated at 37uC, during 1 h. The reaction was stopped by the addition of 80 ml of NaOH 0.5 M and the reaction product, p-nitrophenolate, was measured spectrophotometrically at 405 nm, in a microplate reader (Multiskan ASCENT ThermoH).
b -hexosaminidase inhibitory activity. Beyond avoiding bhexosaminidase release, individual naphthoquinones may directly inhibit b-hexosaminidase enzymatic activity. For this, the inhibition of b-hexosaminidase enzymatic activity by naphthoquinones and quercetin was evaluated in an assay similar to the one described above: individual naphthoquinones (5 ml) were incubated with supernatant of degranulated cells where b-hexosaminidase is present (25 ml of supernatant of cells treated with A23187), in presence of 50 ml of substrate solution, during 1 h, at 37uC. The determination was made at 405 nm, in a microplate reader (Multiskan ASCENT ThermoH) [22]. Released histamine quantification. 100 ml of NaOH 1 M and 25 ml of o-phthalaldehyde (OPA) 1% (w/v) were added to 500 ml of supernatant to convert histamine into fluorescent histamine-OPA-products. After 4 min incubation at room temperature, the reaction was stopped with of 50 ml of HCl 3 M. Precipitated proteins were removed by centrifugation at 14,000 rpm, during 3 min. The fluorescent histamine-OPAproducts were quantified in the supernatant using 360 nm excitation and 450 nm emission in a microplate reader (Biotek Synergy HTH) [21]. Changes in histamine release are expressed as the difference between maximal and basal release, in percentage of control.
Leukotriene C 4 quantification. Leukotriene C 4 quantification was performed in the supernatant of IgE/antigen stimulated cells using a competitive enzyme immunoassay kit, according to the supplier's protocol (Abcam, Cambridge, United Kingdom) in a microplate reader (Multiskan ASCENT ThermoH) at 405 nm.

Assays of enzymatic inhibition in cell-free systems
Hyaluronidase. The enzymatic reaction mixture was composed by 50 ml hyaluronic acid (5 mg/ml in water), 100 ml buffer pH 3.68 (HCOONa 0.2 M, NaCl 0.1 M and BSA 0.2 mg/mL), 200 ml water, 50 ml individual naphthoquinones solution and 50 ml hyaluronidase 600 U. The enzymatic reaction occurred during 1 h, at 37uC. The reaction product, N-acetyl-sugar, was quantified according Morgan-Elson colour reaction with minor modifica-tions. The Morgan-Elson reaction was started by addition of 25 ml disodium tetraborate 0.8 M and subsequent heating in a boiling water bath during 3 min. After cooling, 750 ml p-dimethylaminobenzaldehyde (DMAB) was added and the reaction mixture was incubated at 37uC for 20 min. DMAB stock solution was prepared by dissolving 2 g DMAB in glacial acetic acid with 12.5% of HCl 10 N. This solution was further diluted in glacial acetic acid (1:2) immediately before use. The measurement was made spectrophotometrically, at 560 nm, in a microplate reader (Multiskan ASCENT ThermoH) [23,24]. Sodium cromoglycate was used as a positive control for inhibition [25]. DMSO was kept constant at 1%, without inducing significant enzyme inhibition.
Lipoxidase. Lipoxidase catalyses the oxidation of linoleic acid to the conjugated diene, 13-hydroperoxy linoleic acid, which was measured spectrophotometrically at 234 nm on a UV/visible spectrophotometer (UNICAM Helios a) [26]. The blank was measured in a reaction mixture with 20 ml of individual naphthoquinones solution, 1 ml of phosphate buffer (pH 9) and 20 ml of soybean lipoxidase 500 U. After 5 min pre-incubation at room temperature, the reaction was started by addition of 50 ml of substrate (linoleic acid) at 2 mM in ethanol. The reaction time was 3 min. DMSO was kept constant at 1.8%, without inducing enzyme inhibition. Quercetin was used as positive control [27,28].

Naphthazarin and diospyrin decreased RBL-2H3 degranulation
To test the anti-allergic properties of naphthoquinones, we evaluated their ability to inhibit RBL-2H3 basophils' degranulation evoked by two different stimuli: IgE/antigen (100 ng/mL, 16 h exposure) or the calcium ionophore A23187 (1 mM, 30 min exposure); Schematic protocols in Fig. 2. DMSO was used as a solvent, and we started by determining the maximal DMSO concentrations that could be used without interfering with the assays, and consequently the maximal concentrations that could be tested for the dissolved naphthoquinones (Fig. 3). DMSO at 0.5% decreased b-hexosaminidase release by 52.1611.9% when IgE/ antigen was used (P,0.05) (Fig. 3A), but it had no detectable effect on the release of b-hexosaminidase and histamine in A23187 stimulated cells (Fig. 3B), likely due to the shorter exposure time. Thus, the DMSO amount used in the IgE/antigen assay was 0.1%, while in the A23187 assay it was 0.5%, thus allowing for higher naphthoquinone concentration testing. None of the naphthoquinone concentrations used significantly affected cell viability, as assessed by the MTT reduction assay ( Fig. 4; black bars).
In the A23187 assay, the positive control quercetin required a 10-fold higher concentration (100 mM) to reduce b-hexosaminidase release, when compared with the IgE/antigen assay. Diospirin (DPR) at the higher 10 mM concentration allowed by this assay, significantly decreased both b-hexosaminidase   (56.8614.6%) and histamine (51.4612.8%) release, an amplitude of effect approaching that achieved with quercetin, and standing out amongst the other dimeric and monomeric naphthoquinones, none of which significantly affected degranulation at the maximal tested concentrations (Fig. 4B). None of the tested naphthoquinone concentrations induced degranulation in the absence of stimuli (IgE/Antigen or A23187) nor directly inhibited the b-hexosaminidase enzymatic activity (data not shown).

Naphthoquinones are weak hyaluronidase inhibitors
Sodium cromoglycate was used as a positive control for hyaluronidase inhibition [25]. Complete hyaluronidase inhibition required 10 mM sodium cromoglycate (Fig. 5). DMSO amount precluded the testing of naphthoquinone concentrations above 100 mM. Only menadione and naphthazarin significantly inhibited hyaluronidase: at 100 mM, menadione and naphthazarin inhibited 23.866.06% and 16.962.45% the activity of the enzyme, respectively. Despite their modest inhibition of hyaluronidase, the level of inhibition induced by menadione and naphthazarin at 100 mM surpasses that achieved with the same sodium cromoglycate concentration (Fig. 5).

Naphthoquinones effects on soybean lipoxidase and in leukotriene levels
All tested naphthoquinones and quercetin concentrationdependently inhibited soybean lipoxidase (Fig. 6A and 6B). Dimeric naphthoquinones, diospyrin and diosquinone, were the most potent with respective IC 50 values of 28.9 and 83.8 mM, whereas all monomeric naphthoquinones displayed IC 50 values above 100 mM, with the most potent being menadione with an IC 50 of 128 mM (Table 1). These three most potent naphthoquinones were selected for an exploratory assay on leukotriene levels in IgE/antigen-stimulated RBL-2H3 cells. IgE/antigen treatment induced a robust increase in the levels of leukotriene (LT) C 4 in the supernatant of RBL-2H3 cells, which was unaffected by either diospyrin (1 mM) or diosquinone (1 mM), but completely abrogated by menadione (5 mM). We thus performed a concentration response curve for menadione on LTC 4 levels (Fig. 6C). Results showed that the lowest menadione concentration tested (5 nM) strongly decreased LTC 4 levels evoked by IgE/antigen, suggesting interference with LTC 4 synthesis mechanisms. Increasing menadione concentration-dependently decreased LTC 4 levels until full inhibition of the IgE/antigen evoked release, with an IC 50 of 0.3460.018 mM (Fig. 6C).

Discussion
In this work we investigated the anti-allergic potential of monomeric and dimeric naphthoquinones (Fig. 1) by testing for inhibition of RBL-2H3 cells' degranulation. RBL-2H3 cells are a rat basophilic leukaemia cell line, expressing high affinity IgE receptors (FceRI), being a model to study allergy and inflammation [29,30]. Potent inflammatory mediators (histamine, proteases, cytokines, arachidonic acid metabolites and chemotactic factors) are released from immune cells after an allergic stimulus that can be IgE-dependent or IgE-independent [31]. To induce degranulation we used two previously described effective degranulation stimuli for RBL-2H3 [21]: IgE/antigen (simulation of IgEdependent allergic response) and calcium ionophore (A23187; simulation of events that immediately precede degranulation: increase of intracellular calcium) (Fig. 7). These complementary stimuli assist characterization of the mechanisms by which the studied compounds reduce degranulation. In the present study, the release of immune cell degranulation markers -b-hexosaminidase and histamine [7] -was higher with ionophore than with IgE/ antigen treatment, consistently with previous studies [32].

Naphthoquinone Degranulation inhibition (%) Hyaluronidase inhibition (%) Lipoxidase inhibition (%)
zarin displays the highest structural similarities to shikonin, among the naphthoquinones in the present study. In fact, both compounds share a 5,8-dihydroxy-1,4-naphthoquinone core (Fig. 1). Thus, we propose that naphthazarin and shikonin act through a similar mechanism and that both C5 and C8 hydroxyls modulate direct enzyme interaction via hydrogen bonds [34]. However, other mechanisms of action, that need clarification, could explain the degranulation inhibition verified in presence of naphthazarin, such as a potential binding of naphthazarin to FceRI or to IgE. Diospyrin (10 mM) reduced degranulation in calcium ionophore-stimulated cells (Fig. 4B). Another naphthoquinone, acetylshikonin, was reported to attenuate ionophore-mediated intracellular calcium elevation in rat neutrophils [35]. While, attenuation of calcium elevation might partly explain the effects of both diospyrin and acetylshikonin, their spectrum of activity does not necessarily overlap, since acetylshikonin is reported to decrease leukotriene B4 and tromboxane A2 [35], whereas in the present study diospyrin was unable to reduce leukotrien C4, albeit in different cell models and stimuli. As diospyrin only reduced the A23187-induced degranulation, other common mechanisms of action beyond attenuation of intracellular calcium increase could be excluded, because if diospyrin acted at a level after the intracellular calcium increase, such as the SNARE (soluble Nethylmaleimide-sensitive fusion factor attachment protein receptor) complex formation, diospyrin should have also inhibited the degranulation induced by IgE/antigen complex. Nevertheless, the range of tested concentrations was different. Figure 7. Simplified scheme of RBL-2H3 cells' degranulation pathways. The DNP antigen activates multiple signal transduction pathways via the IgE anti-DNP/FceRI receptor complex. DNP receptor binding activates the immunoreceptor tyrosine activation motifs (ITAM)-Spleen tyrosine kinase (SyK) pathway that can be inhibited by shikonin [35] and probably by naphthazarin. Activated Syk catalyses protein phosphorylation of several proteins, leading indirectly to the activation of protein kinase C (PKC) that induces degranulation and the activation of phospholipase A2 (PLA2). PLA2 increases arachidonic acid (AA) bioavailability that can be converted in leukotrienes (LT) by 5-lipoxygenase (5LO; inhibited by menadione), or in oxidized lipids by means of ROS production. 5LO converts AA into 5-hydroperoxyeicosatetraenoic acid (5-HPETE), which is metabolised to an unstable epoxide, LTA 4 , and finally in LTC 4 , in RBL-2H3 cells. The increase in intracellular calcium by SyK pathway, as well as by A23187 promotes degranulation. doi:10.1371/journal.pone.0090122.g007 Plumbagin was previously reported to exhibit anti-allergic properties, namely, at 5 mM plumbagin inhibited cytokines production by phytohemagglutinin-stimulated peripheral blood mononuclear cells (PBMC) [18]. In the present study, the maximum non-toxic concentration of plumbagin that could be tested in RBL-2H3 cells was 1 mM, and at that concentration plumbagin did not prevent degranulation evoked by either IgE/ antigen or A23187.
Naphthazarin and diospyrin, shown capable of inhibiting RBL-2H3 basophils' degranulation in the present study, warrant further investigation in models such as primary mast cells. Several questions about the membrane permeability of naphthoquinones may be drawn; however, the effects of both naphthazarin (the most hydrophilic), and diospyrin (the most lipophilic) on RBL-2H3 degranulation strongly suggest that naphthoquinones can cross cell membranes. Menadione in particular is well known for its mitochondrial effects [36], and that requires crossing the plasma membrane. Also atovaquone, a monomeric naphthoquinone has oral bioavailability for the treatment and prevention of malaria in humans [37].
Following degranulation studies, we addressed the inhibition of enzymes involved in allergic responses: hyaluronidase and lipoxidase. Hyaluronidase increases vascular permeability in inflammation, by cleavage of internal b-N-acetyl-D-glucosaminidic linkages of hyaluronic acid [38], thus being a possible target for anti-allergic drugs. Our data shows that the tested monomeric and dimeric naphthoquinones are poor hyaluronidase inhibitors, with only menadione and naphthazarin displaying modest inhibitory activity (Fig. 5). 5-Lipoxygenase is a rate-limiting enzyme for leukotriene synthesis, converting arachidonic acid into 5-hydroperoxyeicosatetraenoic acid (5-HPETE). 5-HPETE is metabolised to an unstable epoxide LTA 4 , which is transformed to LTB 4 or LTC 4 according with the cell type and the enzymes present (Fig. 7) [39]. In RBL-2H3 cells, LTC 4 is the major leukotriene released, while LTD 4 and LTE 4 are not produced [40][41][42]. Soybean lipoxidase is often used to model human 5-, 12-and 15lipoxygenases, given the high catalytic domain similarity between plant and mammalian lipoxygenases [43]. All tested naphthoquinones exhibited lipoxidase inhibiting activity (Fig. 6A). Dimeric naphthoquinones (diospyrin and diosquinone), and the monomeric menadione were the most potent, showing the lowest IC 50 values ( Table 1). Considering that these three naphthoquinones are the most lipophilic of the studied compounds, our results raise the hypothesis that naphthoquinones inhibit lipoxidase by competing with natural lipophilic substrates.
Considering the results obtained with soybean lipoxidase, we tested the inhibition of leukotriene production by diospyrin, diosquinone and menadione. Menadione was the only naphthoquinone able to reduce leukotriene production, achieving full inhibition at 5 mM (Fig. 6C). Higher menadione concentrations (50-200 mM) were previously reported to reduce leukotriene production by inhibiting 5-lipoxygenase translocation to the nuclear membrane [17]. Given that menadione is a known oxidative stress generator [44], and that reactive oxygen species (ROS) may react with arachidonic acid forming oxidized lipids (Fig. 7) [45], decreased LTC 4 with our lower menadione concentrations may stem from decreased arachidonic acid availability. Consitently with our concentration range, menadione was reported to inhibit prostaglandin H 2 (PGH 2 ) synthase via ROS production with an IC 50 of 5 mM [46].
Concluding, we evaluated the anti-allergic properties of monomeric and dimeric naphthoquinones by studying the inhibition of RBL-2H3 basophils' degranulation and LTC 4 production induced by allergic stimuli, as well as by the evaluation of inhibition of enzymes involved in allergic responses (main findings for each compound are summarized in Table 2). To our knowledge, this is the first study addressing the anti-allergic potential of diospyrin, diosquinone, naphthazarin and juglone. Naphthazarin and diospyrin reduced degranulation by different mechanisms of action. Naphthazarin and diospyrin acted, respectively, upstream and downstream of the intracellular calcium increase. In spite of being poor inhibitors of hyaluronidase, naphthoquinones inhibited lipoxidase and menadione reduced leukotriene production. Thus, this work expands the current knowledge on the biological properties of naphthoquinones, highlighting naphthazarin, diospyrin and menadione as potential lead compounds for structural modification in the process of improving and developing novel anti-allergic drugs.